An Image Inpainting Approach Based on the Poisson Equation

Abstract

A novel algorithm is proposed for inpainting digital images based on the Poisson equation. In this method, the image to be repaired is decomposed into the structure image and the texture image, which are processed respectively according to their image characteristics. The texture image is repaired by the exemplar-based method; for the structure image, the Laplacian operator is employed to enhance the structure information, and the Laplacian image is inpainted by the exemplar-based algorithm, followed by a reconstruction based on the Poisson equation. The inpainting result is obtained by adding back these two sub-images. Experiments show the proposed algorithm provides a better visual effect, especially for large area inpainting.